Estimation Method Of Parameters For AR(1) Model With Missing Data

The work about analyzing and processing a mass of data exists commonly in various fields of scientific research and actual production.Along with the social development and the intense competition, scholars already show concern over the issue about how to use scientific statistical analysis methods to make an accurate,immediate analysis about data in the collection and formulate the right strategy. Especially,the work about analyzing and processing the missing data is of great significance.AR(1)Model is one of the statistic models which have strong utility and also exists commonly in practical life.Many scholars carried out a great deal of detailed research about this model from various angles.The present paper deals with the estimation problem for zero mean value AR(1)model with missing data under the condition of joint maximum likelihood.The paper is based on the EM algorithm.By using the iterative method,we have obtained specific calculation steps of the estimation methods and the estimated valus of parameters.At last,we did a lot of simulations of my conclusion and made a comparison between my conclusion and the conclusions of authentic data under the jointmaximum likelihood and condition maximum likelihood and missing data.We know that the conclusion of the estimation method under joint maximum likelihood is the best.